It is estimated that about 50,000-100,000 antibodies are currently available worldwide representing about 5,000-10,000 different genes/proteins. This number is significantly below 400,000 proteins in the proteome, and millions of other “antibodiable” antigens. Furthermore, most antibodies don't work as the users intend, and about 75% of antibodies do not work in all antibody-based applications. A key reason of much fewer antibodies than “antibodiable” antigens is that many antigens are hidden in samples or sample preparations and thus poorly accessible to antibodies. This invention discloses a method of designing and detecting hidden antigens.
Post-translational conjugation of a protein by another protein, polysaccharide, lipid and nucleic acid, or any combination of the above plays a key role virtually in every aspect of cellular functions. A conjugated molecule is either a monomeric single-molecule or a polymeric macromolecule with either a linear or a branched structure. Macromolecules include, but are not limited to, proteins or polypeptides, polysaccharides, adenosine diphosphate (ADP)-ribosyls, fatty acids, polynucleotides, glycosylphosphatidylinositol (GPI) anchors, ubiquitin, small ubiquitin-like modifier (SUMO), neural precursor cell expressed, developmentally down-regulated 8 (NEDD8), interferon-stimulated gene 15 kDa (ISG15), and other ubiquitin-like molecules (UBLs).
Many types of human diseases display abnormal molecular conjugation. For example, abnormal glycosylation occurs in many types of cancers (Mehta and Block, 2008). Ubiquitin-containing conjugates are present virtually in all types of neurodegenerative diseases (Dohm et al., 2008). Telomeric aggregates accumulate in tumor cells (Mai and Garini, 2006). Advanced glycation adducts are found in samples obtained from patients with heart disease and/or diabetes (Thornalley, 2002; Meerwaldt et al., 2008). Disease-specific macromolecule-to-macromolecule conjugates are present in body fluids such as blood serum or cerebrospinal fluid (CSF), but few reliable methods are currently available to detect them. However, most, if not all, macromolecule-to-macromolecule conjugation sites are hidden antigens (see FIG. 4 below). Therefore, antibodies to macromolecule-to-macromolecule conjugation sites are difficult to make and are not currently available for assays of macromolecule-to-macromolecule conjugation sites in all antibody-based applications.
Methods of making antibodies against post-translational modified proteins in the form of a small monomeric molecule, including phosphorylation, acetylation, methylation, and nitrolization, are well established. In comparison to monomeric modification site-specific antibodies, there is no effective method currently available for making polymeric macromolecular conjugation site-specific antibodies. Macromolecular conjugation can be defined as covalent conjugation between two polymeric biomolecules, including, but not limited to, protein glycosylation, lipidation, ADP-ribosylation, ubiquitination, sumoylation, NEDDylation, ISGylation, GPI-anchor, transglutaminase-mediated cross-links, and the like.
In the post-genomic era, our knowledge of macromolecule-to-macromolecule conjugation and its relation to diseases has grown exponentially. This provides an opportunity to develop novel methods for detecting macromolecule-to-macromolecule conjugation in a conjugation site-specific manner. For that reason, investigators have devoted extensive efforts to generation of macromolecule-to-macromolecule conjugation site-specific antibodies by conventional antigen design, antibody-making, and antigen detection methods. However, these efforts have been so far proven futile (Matsumoto et al., 2008). As a result, there are few conjugation site-specific antibodies currently available. Therefore, new methods for detecting hidden antigen/epitopes, including but not limited to macromolecular conjugation-sites and linear hidden antigens, are desperately needed, and can provide useful tools for all antibody- and antigen-based applications.
An example is making ubiquitin-to-protein conjugation site-specific antibodies. Protein ubiquitination involves virtually all protein degradation as well as other biological processes. There are a few previous reports of generation of anti-polyubiquitin antibodies. Pirim (1998) reported an anti-polyubiquitin antibody. However, this antibody does not recognize isopeptide bond-branched ubiquitin-to-ubiquitin conjugation, which are dominant forms of cellular ubiquitin conjugates. Rather this antibody recognizes head-to-tail (c- to n-terminal conjugation) poly-ubiquitins, which represent a very tiny/small fraction of polyubiquitin transiently formed during synthesis of free ubiquitin.
Fujimuro et al. (2005) reported anti-polyubiquitin monoclonal antibodies named as FK1 and FK2. Both FK1 and FK2 antibodies recognize the polyubiquitin chain. However, there are two fundamental differences between making FK1 and FK2 antibodies and the methods described in the claims of the present invention: (i) FK1 and FK2 were made by using regular polyubiquitin antigens; and (ii) FK1 and FK2 cannot recognize specific conjugation sites of ubiquitinated proteins (Fujimuro et al., 2005). Therefore, FK1 and FK2 are not conjugation site-specific antibodies, rather than general polyubiquitin antibodies.
Similarly, there are reports of using antibodies to the glycine-glycine-to-(lysine) structure for profiling of ubiquitinated proteins and identification of the peptides with liquid chromatography-tandem mass spectrometry (LC-MS/MS) (Peng et al., 2003; Denis et al., 2007; Xu et al., 2010). The glycine-glycine-to-(lysine) structure was prepared by reacting lysine-rich histone III-S protein with t-butyloxycarbonyl-Gly-Gly-N-hydroxysuccinimide (Boc-Gly-Gly-NHS) (Xu et al., 2010). However, the glycine-glycine-to-(lysine) antibodies recognize only the glycine-glycine-to-(lysine), rather than the lysine surrounding sequence of a specific ubiquitinated protein, and thus they are also not conjugation site-specific antibodies, and cannot be used to detect individual ubiquitin-to-protein conjugation sites. In addition, the glycine-glycine-to-(lysine) antibodies cannot be used in regular antibody-based applications, rather they were developed for the pre-LC-MS/MS profiling applications (Peng et al., 2003; Denis et al., 2007; Xu et al., 2010). In comparison, the inventive methods are for designing and detecting the specific conjugation sites of both conjugation moieties. Therefore, conjugation site-specific antibodies developed by the inventive ACE methods, can recognize specifically both the branched glycine-glycine and the conjugation site lysine surrounding sequence as well as can be used for detecting hidden antigens in all antibody-based applications (see FIGS. 1-13).
Matsumoto et al. (2008) generated two linkage-specific antibodies that recognize polyubiquitin chains through lysine 63 (K63) or 48 (K48) linkage (US patent 20070218069A). However, there are several fundamental differences between the method of making these two linkage-specific antibodies and the methods of the present invention. The “antibodies” made by Matsumoto et al. (2008) were not generated by conventional animal immunization methods, rather by a phage display approach of random screening of the ubiquitin conjugation site binders. This phage display approach has advantage to be able to select binding partners from millions of other irrelevant proteins, but these binding partners are “antibody-like” fusion proteins. Also, the phage display approach usually has technical challenges associated with it. For instance, it is acknowledged that the affinity and specificity of binding partners generated by phage display are often suboptimal, relative to conventional antibody, and the loss of the original heavy- and light-chain pairings is also a challenge. Perhaps for these reasons, phage display has not been widely used to make “antibodies” (Ward, 2002). In comparison, the present invention uses the Artificially Cleaved Epitope (ACE, see below) strategy for designing and detecting macromolecular conjugation site-specific and linear hidden antigens, which are proven to be more effective and reliable (see FIGS. 1-14 below).
There are several patented methods for making peptide antibodies. Patent WO 02/25287 describes methods for analysis of proteins by producing a mixture of peptides and contacting the mixture of peptides to filtering agents or antibodies in order to decrease the complexity of a mixture prior to the application of an analytical technique such as mass spectrometry. U.S. Pat. No. 7,460,960 described methods by the use of capture agents or antibodies that interact with the Proteome Epitope Tags (PETs) in a sample. However, these methods cannot be used to design and detect hidden antigens, and they are also principally and profoundly different with the methods of the present invention.
Currently, there are several cleavage site-specific antibodies commercially available. U.S. Pat. No. 7,803,553 by Kojima et al. described an antibody for detecting an active form of TGF-β1 naturally cleaved in vivo. U.S. Pat. No. 6,762,045 by Krebs et al. described an antibody to naturally cleaved caspase-3. All currently available cleavage-specific antibodies were developed to detect the naturally occurring cleavage sites in vivo, and cannot be used to detect hidden antigens such as macromolecule-to-macromolecule conjugation sites. In contrast, the present inventive methods are to design and detect “Artificially Cleaved Epitopes (ACEs, see below)” of hidden antigens that are not naturally present or exposed. The inventive ACE methods do not include those for detecting naturally cleaved epitopes in samples.
Macromolecules other than polypeptides can also be used to generate antibodies successfully, including, but not limited to, antibodies to lipids, nucleic acids, and saccharide. For example, a mouse monoclonal antibody (e.g., CTD110.6) recognizing the single O-linked N-acetylglucosamine is commercially available. A mouse antibody (e.g., clone 26-5) to a lipid structure is also reported (Young et al., 1987). However, polysaccharide-to-protein and lipid-to-protein conjugation site-specific antibodies are not currently available, most probably because they are hidden antigen and no reliable methods can successfully make them.